A Whittle Index Policy for Dynamic Task Scheduling of Multi-Platform Multi-Sensor Systems

Task scheduling is key to multi-platform multi-sensor systems, which faces challenges of complex sensor combination and high computational complexity. To deal with these challenges, this research introduces the concept of a sensor resource pool to characterize available sensors, and formulate the optimization problem as a restless multi-armed bandit (RMAB) process, considering both task execution and the limitations of the sensor resource pool. To tackle the formulated RMAB problem, we apply Whittle relaxation. Accordingly, the Whittle index policy is introduced for the dynamic task scheduling of multi-platform multi-sensor systems. Numerical simulations demonstrate the effectiveness of the proposed policy.